jsd

Notes on Teach­ing in Prison

A The­ory of Un­su­per­vised Trans­la­tion Mo­ti­vated by Un­der­stand­ing An­i­mal Communication

“Ac­qui­si­tion of Chess Knowl­edge in AlphaZero”: prob­ing AZ over time

• jsd 5 Oct 2023 17:42 UTC

  10 points

  2

  on: Re­sponse to Quintin Pope’s Evolu­tion Pro­vides No Ev­i­dence For the Sharp Left Turn

  Quinton’s

  His name is Quintin (EDIT: now fixed)

• jsd 22 Apr 2023 0:14 UTC

  10 points

  0

  in reply to: Vitor’s comment on: Notes on Teach­ing in Prison

  How did you end up doing this work? Did you deliberately seek it out?

  I went to a French engineering school which is also a military school. During the first year (which corresponds to junior year in US undergrad), each student typically spends around six months in an armed forces regiment after basic training.

  Students get some amount of choice of where to spend these six months among a list of options, and there are also some opportunities outside of the military: these include working as a teaching assistant in some high schools, working for some charitable organizations, and working as a teacher in prison.

  When the time came for me to choose, I had an injured shoulder, was generally physically weak, and did not have much of a “soldier mindset” (ha). So I chose the option that seemed most interesting among the nonmilitary ones: teaching in prison. Every year around five students in a class of about 500 do the same.

  What are teachers, probation officers and so on (everyone who is not a guard) like? What drives them?

  Overall, I thought the teachers were fairly similar to normal public school teachers—in fact, some of them worked part-time at a normal high school. In the prison I worked at, they seemed maybe more dedicated and open-minded than normal public school teachers, but I don’t remember them super well.

  I don’t really remember my conversations with probation officers. One thing that struck me was that in my prison, maybe 90% of them were women. If I remember correctly, most were in their thirties.

• jsd 23 Feb 2023 6:06 UTC

  10 points

  0

  on: Please don’t throw your mind away

  Thanks for this. I’ve been thinking about what to do, as well as where and with whom to live over the next few years. This post highlights important things missing from default plans.

  It makes me more excited about having independence, space to think, and a close circle of trusted friends (vs being managed /​ managing, anxious about urgent todos, and part of a scene).

  I’ve spent more time thinking about math completely unrelated to my work after reading this post.

  The theoretical justifications are more subtle, and seem closer to true, than previous justifications I’ve seen for related ideas.

  The dialog doesn’t overstate its case and acknowledges some tradeoffs that I think can be real—eg I do think there is some good urgent real thinking going on, that some people are a good fit for it, and can make a reasonable choice to do less serious play.

• jsd 28 Oct 2022 15:32 UTC

  7 points

  0

  on: Six (and a half) in­tu­itions for KL divergence

  Thanks for this post! Relatedly, Simon DeDeo had a thread on different ways the KL-divergence pops up in many fields:

  Kullback-Leibler divergence has an enormous number of interpretations and uses: psychological, epistemic, thermodynamic, statistical, computational, geometrical… I am pretty sure I could teach an entire graduate seminar on it.

  Psychological: an excellent predictor of where attention is directed. http://​​ilab.usc.edu/​​surprise/​​

  Epistemic: a normative measure of where you ought to direct your experimental efforts (maximize expected model-breaking) http://​​www.jstor.org/​​stable/​​4623265

  Thermodynamic: a measure of work you can extract from an out-of-equlibrium system as it relaxes to equilibrium.

  Statistical: too many to count, but (e.g.) a measure of the failure of an approximation method. https://​​www.countbayesie.com/​​blog/​​2017/​​5/​​9/​​kullback-leibler-divergence-explained

  Computational (machine learning): a measure of model inefficiency—the extent to which it retains useless information. https://​​arxiv.org/​​abs/​​1203.3271

  Computational (compression): the extent to which a compression algorithm designed for one system fails when applied to another.

  Geometrical: the (non-metric!) connection when one extends differential geometry to the probability simplex.

  Biological: the extent to which subsystems co-compute.

  Machine learning: the basic loss function for autoencoders, deep learning, etc. (people call it the “cross-entropy”)

  Algorithmic fairness. How to optimally constrain a prediction algorithm when ensuring compliance with laws on equitable treatment. https://​​arxiv.org/​​abs/​​1412.4643

  Cultural evolution: a metric (we believe) for the study of individual exploration and innovation tasks… https://​​www.sciencedirect.com/​​science/​​article/​​pii/​​S0010027716302840 …

  Digital humanism: Kullback-Leibler divergence is related to TFIDF, but with much nicer properties when it comes to coarse-graining. (The most distinctive words have the highest partial-KL when teasing apart documents; stopwords have the lowest) http://​​www.mdpi.com/​​1099-4300/​​15/​​6/​​2246

  Mutual information: Well, it’s a special case of Kullback-Leibler—the extent to which you’re surprised by (arbitrary) correlations between a pair of variables if you believe they’re independent.

  Statistics: it’s the underlying justification for the Akiake Information Criterion, used for model selection.

  Philosophy of mind: It’s the “free energy” term in the predictive brain account of perception and consciousness. See Andy Clark’s new book or https://​​link.springer.com/​​article/​​10.1007%2Fs11229-017-1534-5

• jsd 4 Sep 2021 17:05 UTC

  7 points

  on: 18 pos­si­ble mean­ings of “I Like Red”

  Another meaning could be: I want to raise the salience of the issue ‘Red vs Not Red’, I want to convey that ‘Red vs Not Red’ is an underrated axis. I think this is also an example of level 4?

• jsd 10 Nov 2023 23:44 UTC

  6 points

  3

  in reply to: Daniel Kokotajlo’s comment on: AI Timelines

  As previously discussed a couple times on this website

  For context, Daniel wrote Is this a good way to bet on short timelines? (which I didn’t know about when writing this comment) 3 years ago.

  HT Alex Lawsen for the link.

• jsd 29 Mar 2022 4:21 UTC

  6 points

  in reply to: TLW’s comment on: Some thoughts about de­cep­tive mesaoptimization

  It’s an example first written about by Paul Christiano here (at the beginning of Part III).

  The idea is this: suppose we want to ensure that our model has acceptable behavior even in worst-case situations. One idea would be to do adversarial training: at every step during training, train an adversary model to find inputs on which the model behaves unacceptably, and penalize the model accordingly.

  If the adversary is able to uncover all the worst-case inputs, this penalization ensures we end up with a model that behaves acceptably on all inputs.

  RSA-2048 is a somewhat contrived but illustrative example of how this strategy could fail:

  As a simple but silly example, suppose our model works as follows:

  • Pick a cryptographic puzzle (e.g. “factor RSA-2048”).

  • When it sees a solution to that puzzle, it behaves badly.

  • Otherwise, it behaves well.

  Even the adversary understands perfectly what this model is doing, they can’t find an input on which it will behave badly unless they can factor RSA-2048. But if deployed in the world, this model will eventually behave badly.

  In particular:

  Even the adversary understands perfectly what this model is doing, they can’t find an input on which it will behave badly unless they can factor RSA-2048

  This is because during training, as is the case now, we and the adversary we build are unable to factor RSA-2048.

  But if deployed in the world, this model will eventually behave badly.

  This is because (or assumes that) at some point in the future, a factorization of RSA-2048 will exist and become available.

• jsd 3 Nov 2023 21:54 UTC

  LW: 5 AF: 2

  3

  AF

  on: Thoughts on open source AI

  If they are right then this protocol boils down to “evaluate, then open source.” I think there are advantages to having a policy which specializes to what AI safety folks want if AI safety folks are correct about the future and specializes to what open source folks want if open source folks are correct about the future.

  In practice, arguing that your evaluations show open-sourcing is safe may involve a bunch of paperwork and maybe lawyer fees. If so, this would be a big barrier for small teams, so I expect open-source advocates not to be happy with such a trajectory.

• jsd 6 Oct 2023 3:41 UTC

  5 points

  0

  in reply to: aysja’s comment on: Thomas Kwa’s MIRI re­search experience

  I don’t have the energy to contribute actual thoughts, but here are a few links that may be relevant to this conversation:

  • Sequencing is the new microscope, by Laura Deming

  • On whether neuroscience is primarily data, rather than theory, bottlenecked:

    • Could a neuroscientist understand a microprocessor?, by Eric Jonas

    • This footnote on computational neuroscience, by Jascha Sohl-Dickstein

• jsd 21 Oct 2021 16:13 UTC

  5 points

  on: Emer­gent mod­u­lar­ity and safety

  On one hand, Olah et al.’s (2020) investigations find circuits which implement human-comprehensible functions.

  At a higher level, they also find that different branches (when the modularity is enforced already by the architecture) tend to contain different features.

• jsd 20 Apr 2023 20:09 UTC

  4 points

  1

  in reply to: M. Y. Zuo’s comment on: Notes on Teach­ing in Prison

  How is it logistically possible for the guards to go on strike?

  Who was doing all the routine work of operating cell doors, cameras, and other security facilities?

  This is a good question. In France, prison guards are not allowed to strike (like most police, military, and judges). At the time, the penitentiary administration asked for sanctions against guards who took part in the strike, but I think most were not applied because there was a shortage of guards.

  In practice, guards were replaced by gendarmes, and work was reduced to the basics (e.g. security and food). In particular, classes, visits, and yard time were greatly reduced and sometimes completely cut, which heightened tensions with inmates.

• jsd 13 Apr 2022 17:12 UTC

  4 points

  on: jsd’s Shortform

  I’ve been thinking about these two quotes from AXRP a lot lately:

  From Richard Ngo’s interview:

  Richard Ngo: Probably the main answer is just the thing I was saying before about how we want to be clear about where the work is being done in a specific alignment proposal. And it seems important to think about having something that doesn’t just shuffle the optimization pressure around, but really gives us some deeper reason to think that the problem is being solved. One example is when it comes to Paul Christiano’s work on amplification, I think one core insight that’s doing a lot of the work is that imitation can be very powerful without being equivalently dangerous. So yeah, this idea that instead of optimizing for a target, you can just optimize to be similar to humans, and that might still get you a very long way. And then another related insight that makes amplification promising is the idea that decomposing tasks can leverage human abilities in a powerful way.

  Richard Ngo: Now, I don’t think that those are anywhere near complete ways of addressing the problem, but they gesture towards where the work is being done. Whereas for some other proposals, I don’t think there’s an equivalent story about what’s the deeper idea or principle that’s allowing the work to be done to solve this difficult problem.

  From Paul Christiano’s interview:

  Paul Christiano: And it’s nice to have a problem statement which is entirely external to the algorithm. If you want to just say, “here’s the assumption we’re making now; I want to solve that problem”, it’s great to have an assumption on the environment be your assumption. There’re some risk if you say, “Oh, our assumption is going to be that the agent’s going to internalize whatever objective we use to train it.” The definition of that assumption is stated in terms of, it’s kind of like helping yourself to some sort of magical ingredient. And, if you optimize for solving that problem, you’re going to push into a part of the space where that magical ingredient was doing a really large part of the work. Which I think is a much more dangerous dynamic. If the assumption is just on the environment, in some sense, you’re limited in how much of that you can do. You have to solve the remaining part of the problem you didn’t assume away. And I’m really scared of sub-problems which just assume that some part of the algorithm will work well, because I think you often just end up pushing an inordinate amount of the difficulty into that step.

• jsd 12 Jan 2022 10:22 UTC

  4 points

  on: The Best Soft­ware For Every Need

  Software: streamlit.io

  Need: making small webapps to display or visualize results

  Other programs I’ve tried: R shiny, ipywidgets

  I find streamlit extremely simple to use, it interoperates well with other libraries (eg pandas or matplotlib), the webapps render well and are easy to share, either temporarily through ngrok, or with https://​​share.streamlit.io/​​.

• jsd 21 Oct 2021 16:18 UTC

  LW: 4 AF: 3

  AF

  in reply to: Jsevillamol’s comment on: Emer­gent mod­u­lar­ity and safety

  Relevant related work : NNs are surprisingly modular

  https://​​arxiv.org/​​abs/​​2003.04881v2?ref=mlnew

  I believe Richard linked to Clusterability in Neural Networks, which has superseded Pruned Neural Networks are Surprisingly Modular.

  The same authors also recently published Detecting Modularity in Deep Neural Networks.

• jsd 1 Apr 2024 12:45 UTC

  3 points

  0

  on: The Best Tacit Knowl­edge Videos on Every Subject

  Related: Film Study for Research

• jsd 17 Nov 2023 17:45 UTC

  3 points

  0

  on: Me­mory band­width con­straints im­ply economies of scale in AI inference

  According to SemiAnalysis in July:

  OpenAI regularly hits a batch size of 4k+ on their inference clusters, which means even with optimal load balancing between experts, the experts only have batch sizes of ~500. This requires very large amounts of usage to achieve.

  Our understanding is that OpenAI runs inference on a cluster of 128 GPUs. They have multiple of these clusters in multiple datacenters and geographies. The inference is done at 8-way tensor parallelism and 16-way pipeline parallelism. Each node of 8 GPUs has only ~130B parameters, or less than 30GB per GPU at FP16 and less than 15GB at FP8/​int8. This enables inference to be run on 40GB A100’s as long as the KV cache size across all batches doesn’t balloon too large.

• jsd 17 Nov 2023 2:45 UTC

  3 points

  1

  on: The 6D effect: When com­pa­nies take risks, one email can be very pow­er­ful.

  I’m grateful for this post: it gives simple concrete advice that I intend to follow, and that I hadn’t thought of. Thanks.